Design of stacked halo antenna for 862 MHz LTE

[jonasolof]

Stacked halo antennas have been used by some radio amateurs who wish to have horisontally polarized UFH antennas.

In certain Europen countries, i.e. Germany and Sweden, LTE 800 has a center frequency of about 850 MH for the user up link. If you are in radio shadow and therefore depend on reflected signals, you'd rather use an omnidirectional antenna. At the same time, routers have two antenna connectors to allow for polarization diversity. My idea is to use a good vertical antenna such as a Procomm.dk marine/basestation antenna for vertical polarization and a multiple stacked halo antenna for horizontal polarization. Four stacked halos should give about 6dB gain. However, I haven't found any info on the vertical separation distance of the halos and the length of the matching coaxial cables. I understand that you'd probably need an antenna lab to match the halos, that is available. I would be glad for any info on designing a 4 to 8 stacked multiple halo for 847 MHz.

Or is there any other design for an omnidirectional horizontally polarized antenna with a theoretical gain of 6-8 dBi which could suit the purpose?

Regards

Jonas

 

[VEBill]

Is this for use at a fixed-location (such as a house)? If so, then perhaps a single high gain antenna aimed in the optimum direction (towards the tower with the best signal) would be a better and simpler solution.

 

[jonasolof]

Hi, it isn't for a specific location but out of interest in the design concept. It is easy to find high gain yagis and also double quad panel antennas with 8dB gain combining two quads at + - 45 degrees to give polarization diversity. LTE supports MIMO and router receivers can handle multipath signals which is what one would have at a site where the only signals available are reflected. You might not even have a general direction from which signals come. The question arouse out of a LTE broadband user's conundrum and questions on a Swedish broad band forum. The nearest base station is obscured by a large industrial building which effectively blocks signals. Hence, the only signals which he can receive would be reflected and multipsth. He could ask for a commercial consultant to use a directional scanner to see if there is a main direction for acquiring a good signal but that costs many hundred euros and even more dollars.

I don't think that he can rely on diffraction over the top of the building either. In this situation a combined vertically and horizontally polarized omnidirectional antenna arrangement would be of interest.

Regards

Jonas
SM 0 WJY

 

[VEBill]

What about the 2nd nearest tower?

These sorts of systems (LTE or anything similar) are typically designed to be used with low gain antennas. At these frequencies it's relatively easy to achieve extremely high gain in a fixed direction - Yagi or corner reflector.

Start with whatever is easiest, before embarking on more complicated solutions.

 

[jonasolof]

Having read a number of articles on stacked halo antennas, I've realized that the SWR increases very rapidly as you deviate from the design frequency. This is a disadvantage. Otherwise, construction of a four-stacked halo seems feasible. A design for 431 MZ had 17" vertical spacing between loops. This would be 21 cm on 862 MZ making the height of the entire assembly 84 cm for four loops.

The specific question for this thread was on the design of a stacked halo, not for the most convenient way to solve a specific problem reating to one location. The user behind the factory building has already got advice on how to use his two logperiodic Poynting antennas and search for best signal, and also eventually to try a vertical omnidirectional antenna for the vertical polarization instead of a directional antenna.

 

[jonasolof]

Could it be that a Gamna match makes the stacked halo antenna more narrowbanded from a SWR point of view than if another match such as a coaxial balun is used?

 

[VEBill]

One of the costs of free advice is that you'll be given advice that you might not want to see. It comes with the territory. We are not 'yes men' on this forum.

Skip over the next two paragraphs if you're really not interested.

As a rule of thumb (I'm sure you know), doubling is +3dB. So four stacked antennas will be +6dB compared to one. But the more complex one-to-four feed network might introduce one or two additional dB loss compared to one antenna with a simple feed. So all this work for four stacked halo antennas is heading towards maybe +4 or +5 dB. Maybe less depending on cable loss at this band. I'd call that "a slight improvement".

If there's no signal at all, then you'll need much more than a slight improvement to make it work reliably. Your approach would be ideal if the current set-up already worked most of the time and just requires a slight improvement to improve the reliability on bad days.

Resume here:

There are a multitude of antenna match designs, some are wider in bandwidth than others. The Ham radio antenna handbooks (ARRL, RSGB) have tonnes of info on this subject.

I don't know the LTE band plan for your location, but it's not likely to be a large percentage bandwidth. In any case, moderate VSWR is not typically a real world issue. The antenna system worries less about moderate VSWR than the humans, and cable loss will mask much of it from the transmitter.

 

[jonasolof]

The LTE plan for 800 Mhz is 60 MHz wide if we consider both up and down link plus the gap in between. For 2600 it is 120 Mhz. For a specific operator the up link is 10 MHz wide at 800 and 20 Mhz wide at 2600.



A combination of a vertical omnidirectional antenna plus an omnidirectional horizontally polarized antenna is obviously not an all purpose antenna. Only in a few scenarios would it be useful, just like the ham operator's stacked two meter halo.

I was a licensed radio amateur already in 1959 and although I wasn't active for long as I went to university, my practical interest in antennas and radio wave propagation has continued. I was one of the first persons in Scandinavia to get a 900 Nordic Mobile Telephone handheld cellphone in 1987. In these early years with scarce coveasge, knowing the importance of antenna height, reflections, diffraction, propagation over water and antenna designs helped me to set up communication when others failed. Later on, i got interested in the propagation apects of the CDMA2000 system used for large cell broadband on 450 MHz. Cell range can be above 120 km, which means that antenna height and gain is of great importance. Today, i have tested a German LTE router in Sweden, a router which hasn't been tested here before but performed extremely well on both 800 and 2600 MHz. For speed, a good signal is needed. For some initial tests, a reasonably good omnidirectional antenna can be useful: Q: Six meters up, what b roadband providers can be picked up, what is the cell id of the particular mast and what is the relative strength of different signals, measured in all of RSRP, RSRQ and SINR. This can help you decide on the permanent antenna solution.

So I know when a couple of decibels matter.

Those that have built four stacked halos for the 2 meter and 70 cm bands have also included the losses in their appraisal of the design. The figures presented are the same as your estimates. They knew this beforehand but still decided to build. And results were OK within the confines of their reasonable expectations.

If you have a weak signal picked up by an omnidirectional vertical antenna, the most effective remedy is often to put it higher. On that pole, one could very well add a side bar with the stacked halo and thus benefit from polarization diversity. Nokia showed in research in the nineties that slanted +-45 degrees polarization at the base station only incurred a liss of -1dB in relation to spatial diversity. This was tested in field meadurements with cell phones held t different angles. For LTE, polaization diversity is part of the idea.


As for undsollicited advice, I chuckle. As a 68 years old doctor, I allow myself to experiment if I don' t harm others. Dealing with people, I try not to overdose unsollicited opinions of mine. It is seldom effective, sometimes because people's ideas are so warped. But then, as the Russian saying goes, the grave straightens out even the hunchback.

I'll check the handbooks and see what input I can get from other forums and newsgrpups.

73

 

[VEBill]

Another approach to polarization diversity is to use a CP antenna. One will lose 3dB gain (linear to CP), but will gain polarization independence. Using a long axial mode helical antenna you can achieve very high gain. Crossed Yagi is another approach.

My previous suggestions are based on an assumption: it would be better, faster and cheaper to use the traditional approach of selecting higher gain antennas. Cheap high gain Yagi antennas are readily available (Google: LTE yagi -TV).

Cable loss could be elminated by remoting the modem and using wifi to bridge the final 100m (or more with another 2.4GHz Yagi). The remote box would need only AC power. Extremely high performance is easily achievable.

Good luck.